Wintsch Labs - Out of Business

[DNA]

Per Frozen CPU, Wintsch Labs went out of business. Their TEC products are no longer available.

Does anyone know the details as to why?

[Lothar5150]

Originally Posted by DNA

Per Frozen CPU, Wintsch Labs went out of business. Their TEC products are no longer available.

Does anyone know the details as to why?

I can make a damn good guess....Their products were geared toward the smallest segment of a niche market.

[ricecrispi]

More like a niche of the smallest nicheniche.
Not just peltier cooling, extreme peltier cooling and with very expensive parts.

[brucoman]

too bad
whilst true about target market, those of us IN that market lose another vendor

[BillA]

unless I mistake this co for another, nothing at all but a ripoff of Swiftech's designs ?
did that co 'contribute' a single new design ?

those of you 'in that market' quite apparently do not understand much about it
by eroding Swiftech's market, investment in new products cannot be justified
you guys are your own worst enemies

[flatline]

show me the swiftech tec powered full cover gpu block pls

[brucoman]

Bill, I do not consider them a ripoff of Swiftech designs, but that is my opinion.

As noted above, Wintsch did create a unique vid TEC as well.

Lastly, you may remember (or not) that I use Swiftech gear but I really don't care if another company has the effect of "eroding Swiftech's market". If a tiny bit player such as Wintsch has such a detrimental impact on Swiftech's bottom line, then perhaps Swiftech should focus their time and money on areas other than TECs.

[BillA]

"tec powered full cover gpu block"
?
you mean inc ram ?
if so, bogus product (no known bebefit to WCing ram)
historically, Swiftech would not offer a product w/o some performance bump
but SLI is huge pressure

the problem is not for Swiftech, but for all mfgrs
there must be the expectation of sufficient volume - or higher prices
fragmentation of the market may kill new product offerings, Swiftech is not profitable, eh ?
you may be back to DD

[flatline]

so its bouth a "ripoff of Swiftech's designs" and something "Swiftech would not offer" at the same time a marvel indeed

[BillA]

clearly not so marvelous as your appraisal of my words

[flatline]

you mean the all knowing word of god?

[pHaestus]

I wasn't aware they made a GPU block either. Their CPU block did bear a stricking resemblance to an MCW-5002T though.

[brucoman]

http://www.frozencpu.com/ex-blc-161.html is the gpu block, works for 6800 series SLI too....need a tweak for 7800

[BillA]

had not seen, if comment above applies
coplanar ? tolerance stack ? whew, my head hurts to see how this is going to be a top performer
at $199 its about bragging

[maxSaleen]

Yeah, their CPU block did look just like the 5002, even the internals looked similar. The difference was that it had 2 inlets and 2 outlets to facilitate a greater flow of liquid through the block. I never thought they'd last. Not too many pelters around anymore.

Eeesh that GPU block is ugly. And a waste too. Why would you want your RAM that cold? RAM isn't ment to run at sub ambient temps. It's almost like chilling your MOSFETs, they will probably function worse.

Survival of the fittest....

[Brians256]

Originally Posted by maxSaleen

Yeah, their CPU block did look just like the 5002, even the internals looked similar. The difference was that it had 2 inlets and 2 outlets to facilitate a greater flow of liquid through the block. I never thought they'd last. Not too many pelters around anymore.

Eeesh that GPU block is ugly. And a waste too. Why would you want your RAM that cold? RAM isn't ment to run at sub ambient temps. It's almost like chilling your MOSFETs, they will probably function worse.

Survival of the fittest....

[Tangent]
OK, chilling your MOSFETs might make them worse? The only reason I can think that would make sense is if the lowered resistance (Rdson) increased noise because the R in an RC noise filter is now less. But, the DC-DC converters usually just use LC filtering, not RC filtering. So, am confused.
[Tangent]

[Bloody_Sorcerer]

Uh, ram does benifit from extreme cold (albeit not much; though colder ram does run tighter timings). someone at XS ran BH-5 under LN2 and got 10-15 more mhz.
also, well... they had a GPU TEC block that didn't suck outright and it took big pelts (50x50, 62x62). Show me the swiftech product that does that, regardless of if it cools the ram or not.

[snowwie]

there are several people who have run their system memory sub-zero and reported higher overclocks (each case involved high voltage memory, the venerable winbond ICs)

The difference was that it had 2 inlets and 2 outlets to facilitate a greater flow of liquid through the block.

how so? pump still has only one outlet and one inlet. pd diff? yeah right, quantify that.

[maxSaleen]

I believe that they suggested running two pumps. Thus it would go:

Rad1->Pump1->Block Inlet 1->Block Oulet 1->Reservior1->Rad1

Rad2->Pump2->Block Inlet 2->Block Oulet 2->Reservior2->Rad2

Or something like that. They suggested this simply to deal with the massive amounts of heat being dumped by their HUGE pelts and the CPU.

I can't explain why MOSFETs don't function properly at low temps, as I'm not an electrical engineer, but I do know this from some of the projects over on the VR forums. One case that comes to mind was when someone wanted to LN2 cool an Opteron when they first came out (Grace maybe? Or was it a Japanese guy? Probably the later). They kept running into problems with board supplying the wrong voltage. It turned out that the MOSFETs had to be insulated in order for the board to function properly at such low temps.

I'd love to see a "substantial" overclock on RAM gained by going sub-ambient. 10mhz doesn't cut it (probably only 5% of actual speed). I could see lower temps helping by reducing the electrical resistance through the RAM, essentially increasing the voltage. There are more effective ways of doing this (volatage mods, etc.).